HAL Id: hal-02897968
https://hal.archives-ouvertes.fr/hal-02897968
Submitted on 18 Dec 2020
HAL is a multi-disciplinary open access
archive for the deposit and dissemination of
sci-entific research documents, whether they are
pub-lished or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diffusion de documents
scientifiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés.
Investigating the properties of humins foams, the porous
carbonaceous materials derived from biorefinery
by-products
Pierluigi Tosi, Gerard P.M. van Klink, Charlotte Hurel, Claire Lomenech,
Alain Celzard, Vanessa Fierro, Clara Delgado-Sanchez, Alice Mija
To cite this version:
Pierluigi Tosi, Gerard P.M. van Klink, Charlotte Hurel, Claire Lomenech, Alain Celzard, et al..
Inves-tigating the properties of humins foams, the porous carbonaceous materials derived from biorefinery
by-products. Applied Materials Today, Elsevier, 2020, 20, pp.100622. �10.1016/j.apmt.2020.100622�.
�hal-02897968�
ContentslistsavailableatScienceDirect
Applied
Materials
Today
jo u r n al hom e p a g e :w w w . e l s e v i e r . c o m / l o c a t e / a p m t
Investigating
the
properties
of
humins
foams,
the
porous
carbonaceous
materials
derived
from
biorefinery
by-products
Pierluigi
Tosi
a,b,
Gerard
P.M.
van
Klink
b,
Charlotte
Hurel
c,
Claire
Lomenech
c,
Alain
Celzard
d,
Vanessa
Fierro
d,
Clara
Delgado-Sanchez
d,
Alice
Mija
a,∗aUniversitéCôted’Azur,InstitutdeChimiedeNice,UMRCNRS7272,06108NiceCedex02,France bAvantiumChemicalsB.V.–Zekeringstraat29,1014BVAmsterdam,TheNetherlands
cUniversityCôted’Azur,CNRS,InstituteofPhysicsofNice(INPHYNIUMR7010),06100Nice,France dUniversitédeLorraine,CNRS,IJL,88000Epinal,France
a
r
t
i
c
l
e
i
n
f
o
Articlehistory:
Received2September2019
Receivedinrevisedform10March2020 Accepted10March2020 Keywords: Polymericfoams Biomassvalorization Auto-crosslinking Gasadsorption Surfaceproperties
a
b
s
t
r
a
c
t
Huminsasbiorefineriesby-productcanbeconvertedwithadirectheatingtreatmentintonewrigid porouscarbonmaterialsknownashuminsfoams.Currently,notmanyinformationsaboutthisnew materialareknown.Here,apreparationprotocolintwostepsinvolvingfoamingandcarbonizationis reported,whilethematerialshavebeeninvestigatedintermsofmorphology,elementalcontent,water adsorption–desorption,stabilitytosolventsandhightemperatures,andthermalconductivity.Inorder toevaluatetheirpotentialusesinapplicationssuchaswaterpurification,thepHassociatedtothezero chargepointhasbeenidentified.Foamspreparedunderairventilatedconditionrevealedanextremely negativesurface,whichcanbeusedincationexchangeapplications.Thesurface’sgroupshavebeen iden-tifiedbyBoehmtitration,showingthatthisbehaviorcanbeassociatedtothepresenceofacidmoieties, whilethebasicspeciesresultabsent.ThehuminsfoamshavebeenalsotestedthroughCO2adsorption testsatrealisticoperationconditions,revealingthattheirperformancescancompetewithmaterials reportedinliterature.Finally,activatedcarbonmonolithsfromcarbonizedhuminsfoamsusingCO2 acti-vationweretested,reachingasurfaceareaof1347m2g−1after20minand1482m2g−1after40minof activation.Thesemonolithshavebeencharacterizedintermsofmorphologyandelementalcontent.The resultsprovethatthehuminsfoamsareveryversatilematerials,costeffectiveandeasytoproduce,with promisingpropertiesthatcanbefurthertailoredforforeseenapplications.
©2020TheAuthor(s).PublishedbyElsevierLtd.ThisisanopenaccessarticleundertheCCBYlicense (http://creativecommons.org/licenses/by/4.0/).
1. Introduction
Developing research and markets around by-products has becomeofhighinterestwithregardstothecirculareconomyof industrialprocessesalongwiththeenvironmentalpointofview. Oftenalsocalled“industrialwaste”,wecanidentifyasby-products anykindofmaterialapartfromtheprimaryproductsforwhich theplantwasoriginallyintended.Inthiscontext,humins consti-tuteoneofthemostpromisingby-productscurrentlyderivedfrom thebiorefineryduringtheconversionoflignocellulosicbiomass. Huminsareproducedduringacid-catalyzeddehydration(ACD)of C6sugarsinintermediatessuchas5-hydroxymethylfurfural(HMF), alkylderivatives(suchas5-alkoxymethylfuraldehydes:RMF),and levulinic acid (LA). Humins consist in a dark-colored material
∗ Correspondingauthor.
E-mailaddress:Alice.MIJA@unice.fr(A.Mija).
derivedfromrandomcondensationsbetweenseveral intermedi-ates (mainlyHMF)during theACDprocess [1–3]. Thechemical structure of humins is mainlybased on a complex network of furanicringsandaliphaticchainsbearingseveralreactive oxygen-based functional groups (hydroxyls, ketones, aldehydes, esters, etc.)[4–6].Duetotheirpoorsolubilityandtoseveralengineering drawbacksintheirproductiononanindustrialscale,the produc-tionofhuminswasconsideredachallengetobeovercome[9–11]. Countlessthenumberofstudiesaimingtoavoidthetheirformation
[1,8,10–22].Despiterecentresearchesreachedinterestinggoalsin reducingthehuminsformationtheseprotocolsaresofarlimitedto lab-scalemethods,andunlikelytobeadoptedinacommercialscale plantbecauseoftechnicalissues,highercosts,yieldsandrecovery challenge[21,23–29].Therefore,asmatteroffact,humins produc-tioniscurrentlyunavoidableatindustriallevel.Evenifanindustrial efficientmethodwillbeintroduced,itwillrequirefurtherdecades tobefullyimplementedinalltheproductioncompartments.On theotherhand,themarketofsustainablederivedproductsisrising, https://doi.org/10.1016/j.apmt.2020.100622
andwithitalsothehuminsproduct.Wecannotaffordtoconsider huminsasbarewaste,especiallyconsideringthatrecentresearches haveprovedthemanyhiddenpossibilitiesofthisfuran-rich mix-ture,whichmightinturnbeattractivefortheinterestingproperties thathuminscanoffer[30–39].
Recently,wehavereportedthepreparationofanewpolymeric macroporousrigidmaterialcalledhuminsfoams[40],whichcan beobtaineddirectlyfromamixtureofindustrialhuminsbya sim-pleone-stepthermalprocess,withoutanykindofmodificationor pre-treatment[41].Bycontrollingthepreparationparameters,it ispossibletoobtainauniformand homogeneousporositywith adjustable cell diameters of between0.2 and 3.6mm, or even porositygradients.Inaddition,closed/opencellsyieldandcarbon contentcan betailoredby choosing theconditions of prepara-tion.Thefoamingmechanismhasbeenidentifiedasacombination ofchemicalreactionsandphysicochemicalprocessestakingplace simultaneously.Thecomplexviscosityofthematerialdropsfrom about2.5×105Pastoaminimumvalueofabout0.48Paswhen heatedfrom20to125◦C.Atthislattertemperature,thematrixis sufficientlyfluidtoallowaneasyevolutionofvolatilesubstances. Fromatemperatureofabout140◦C,severalgases,mainly low-molecularweight species produced during the heating process (e.g.H2O,CH3OH,CO,CO2)arereleased,producingbubblesinthe moltenhuminsmatrix.Thehardeningprocessoccursabove170◦C, resultinginthefinalporosityinthethermosethumins-based mate-rial.
However,itshouldbeemphasizedthatthecompositionofcrude huminsishighlydependentfromtheindustrialprocessinwhich theyareformed,andingeneraltheycanappearinsolidor vis-cousform.Thesolidmaterialisformedwhenhuminsarecollected fullycrosslinked,andinthisformthesearebarelyreactive[2].The processofhuminsfoamsproductioncanbeappliedbyonlyusing huminsnotfullycrosslinked,intheformofhighlyviscousmixture ofoligomers[42,43].Alsointhiscasecrudehuminscomposition’s canvarydependingonthespecificparametersusedintheACDof sugars.Thekeystepsofthefoamingmechanismcanbesubjected tosmalltemperaturedifferencesdependingonthesampleused, andinparticularitsviscosity.However,allthesamplestestedgave verysimilarresults.Thus,ingeneralterms,itcanbeassumedthat thismechanismrepresentswellthehuminsbehaviorsduringthe foamformationprocess.
Given the simple and economically attractive approach of huminsfoamsproduction,aswellasthecurrentmarketofporous polymers and carbon-based materials, this route might repre-sentapromisingvalorizationofhuminsby-products.Inorderto findapplicationsfortheseporousmaterials,their physicochemi-calbehavior,composition,thermalstability,thermalconductivity, surface chemistry and morphology should be studied in more detail.Inthiswork,wepresenttheresultsofaseriesof charac-terizationsthatwillhelpfuturestudiestopromotehuminsfoams andoptimizetheirparametersaccordingtotheintended applica-tion.Furthermore,apreliminarytestofpreparation ofactivated carbonmonolithsfromhuminsfoamshasbeencarriedout. Acti-vatedcarbonpreparationisagenerallyinexpensivewaytoproduce valuablematerialsfromby-productsandwaste[44,45].In2018, Kangetal.[46]reportedthepreparationofactivatedcarbonfrom huminsusingchemicalactivationwithKOHinatemperaturerange of500–900◦C, reachingaBET surfaceareaofbetween428and 1975m2g−1. Unfortunately, thematerial yields werequite low (15–39.2%).Inthesameyear,Chernyshevaetal.[47]reportedthe preparationofactivatedcarbonfromhuminsusingKOH activa-tionandCO2physicalactivation,butreachingaBETareaofonly 862m2g−1inthebestcase.Wedecidedtotestthepreparationof activatedcarbonnotdirectlyfromhuminsbutfromhuminsfoams carbonizedat900◦C,inordertousetheporousrigidcarbon-based structureasawholeandtoproducemonoliths.Unlikethe
conven-tionalpowderforminwhichactivatedcarbonisgenerallyavailable, monolithscanbeeasilyrecoveredandreusedandcantherefore coverawiderrangeofapplications.Herein,wehaveprovedthat itispossibletopreparematerialswithhighsurfaceareathatcan competewiththoseavailablecommercially.
2. Methods
2.1. Crudematerials
Humins were provided by the Avantium N.V pilot plant in Geleen (Netherlands) during the ACD of fructose into methoxymethylfurfural(MMF).Thisprocessisthekeystepinthe productionof2,5-furandicarboxylicacid,intermediateforthe pro-ductionofpolyethylene-furanoate(PEF).
2.2. Preparationofhuminsfoams
Thehuminsfoamssamplesweredesignatedbytheletter“F” fol-lowedbythepreparationtemperature(e.g.,F250waspreparedat 250◦C).Foamstermed“F2”werepreparedwithatwo-stepprotocol (e.g.,F2500waspreparedbycarbonizingF250at500◦C).
SamplesF180,F220,F250andF300werepreparedasfollows: about25gofcrudehuminswereinstalledina rectangular flat-bottom alumina crucible and then transferred to a pre-heated NaberthermL9/11/SKMovenfor1hunderairattemperaturesof 180,220,250or300◦C,respectively.Then,theovenwassetat20◦C, andthesampleswereallowedtocoolslowlydowninsideit(about 2hrequiredtoreachroomtemperature).
F2500,F2700andF2900werepreparedbycarbonizationofF250 inaCarboliteGHC12/900tubefurnace(N2 flow∼80mLmin−1) with the following heat treatment: heating up at 1◦Cmin−1; isotherm1hatthefinaltemperature(500,700and900◦C, respec-tively);then1◦Cmin−1whilecoolingdowntoroomtemperature. 2.3. Elementalanalysis
TheelementalanalysiswascarriedoutwithanElementarVario ELCubeanalyzer.Carbon,hydrogen,nitrogenandsulfurcontents were first determined by combustion of the samples at about 1700◦C(atemperatureinducedina furnace,heatedat1150◦C, byatinfoilwrappingthesamplesandusedascatalyst)inastream ofmixedoxygenandhelium,thelatterbeingusedasacarriergas. Oxygenwasquantifiedwiththesameequipmentinasecondstep, usingadifferentprocedureandanotheranalyticcolumn.
Headspaceanalysis– Headspaceanalyseswere performedin anAgilentJ&WDB624(20m×0.18mm×1m)GC–MSequipped withheadspace-auto-samplerandamass-selectivedetector(MSD) withionizationviaEI+.10mgofhumins orhuminsfoamswere weighed in a 20mL vialand the vialwas sealed.The vial was heatedandshakenfor2h(71shakesmin−1 withanacceleration of260cms−2)intheheadspaceauto-sampler,andasmallamount (∼1mL)ofthevaporwasanalyzedwithGC–MS(inlettemperature 250◦C).TheNISTlibrary11wasusedforidentification.
Densityevaluation–Huminsfoamssampleswerecutinto par-allelepipeds and theirvolume determined accurately. The bulk densityofeachsamplewasevaluatedasweight/volume(gcm−3) ratio,andtheaverageofthreerepetitionswascalculated.
2.4. Masstitration
Threesolutionsof0.2MKClwerepreparedandtheirpHwas adjustedtoeither3,6or9using0.1MNaOH/HClsolutions.The pHwascheckedateverystageusingaMettlerToledoLEpH-meter withLE438-IP67pHelectrode(calibratedevery24h).Eachofthe
aforementionedsolutionswasusedtoprepare3sets of6 solu-tions(20mLeach):6solutionsatpH3,6solutionsatpH6,and 6solutionsatpH9.Ineachsolutionoftheseseries,anincreasing amountofmilledfoamwasadded,sothateachseriescontained 0.05,0.1,0.5,1,5and10wt.%ofsample,respectively.These sus-pensionswerestirredandtheirpHwasmeasuredafter24h.The finalpHofeachsolutionwasplottedagainsttheaddedmassof huminsfoam.ThepHPZCcanbeidentifiedasthepHtowhicheach seriestendswhenthesamplemassincreasestoinfinity.Theresults havebeenreportedastheaverageofthreerepetitions.
2.5. pHdriftmethod
Eightsolutionsof0.02MNaCl(13mLeach)werepreparedby adjustingthepHintherange2–9witheither0.1MNaOHor0.01M HClsolutions.ThepHwascheckedateachstepusingaMettler ToledoLEpH-meterwithLE438-IP67electrode(calibratedevery 24h).Toeachsolution,0.1gofthemilledfoamtobestudiedwas added,andthemixturewasstirredfor24h.ThepHshiftwas mea-suredattheendofthe24h,andtheresultswereplottedasinitial pH(pHin)vsfinalpH(pHfin).ThepHPZCwasidentifiedbythepoint wheretheplottedcurvecrossedthelinepHin=pHfin.Theresults werereportedastheaverageofthreerepetitions.
2.6. Boehmtitration
Thenumberofsurfaceoxygen-containinggroupswascalculated accordingtotheBoehmtitrationmethod[48,49],usingthesame automatictitrationequipmentdescribedintheprevious subsec-tion.20mLofsolutionsof0.05MNaOH,0.05MNaHCO3,0.05M Na2CO3and0.05MHClwereprepared.Toeachsolution,0.4gof milledhuminsfoamswereadded.Theflasksweresealedandthe mixturesstirredwithamagneticbarfor24h.Duringthistime,part ofthehuminssurfacegroupswasneutralizedbytherespective solutions.Thefollowing assumptionswereconsidered: NaHCO3 neutralizes carboxyl groups;Na2CO3 neutralizes carbonyls and lactones;NaOHneutralizescarboxyl/lactone/hydroxyl(phenolic) groups;HClneutralizesallbasicgroups.Thesolutionswerethen filteredtoremovethesolidmaterial.Inordertoknowtheamount ofNaOHandHClneutralizedbythesurfacegroupsofthehumins foam,2mLofthefilteredsolutionsweretransferredto10mLflasks, andtitratedwith0.1MHClandNaOHsolutionsrespectively.Onthe otherhand,2mLoftheNaHCO3/Na2CO3 filteredsolutionswere transferredtoa10mLflask;toeachsolution,2mLof0.1MHCl wereadded,and thetwo resultingsolutionsweretitrated back with0.1MNaOH.Inthiscase,theback-titrationgivesresultsthat aremoreaccurate. Theresultswerecheckedtwicebyasecond repetition.
2.7. Solubilitytest
Inordertoidentifythespeciesreleasedintowater,10gofF180, F250andF300huminsfoamswerethoroughlywashedwith boil-ingdistilledwater(550mL)onaBüchnerfunnelequippedwith filterpaperundervacuumpumping.Thefilteredsolutionswere collected,diluted1/3withdistilledwaterandsaccharin(dissolved inacetonitrile(ACN))addedasaninternalstandard(0.2mgmL−1). ThesolutionswereanalyzedusingaWatersAcquityUPLCHSSC18 columnequippedwithaUVdetector,using0.2%trifluoroaceticacid inH2Oaseluent.Thecolumnwasthermostatedat50◦C.
Thesolubilitytestindifferentsolventswascarriedoutbyadding 0.1gofmilledF250in20mLofACN,EtOH,wateratpH4,andwater atpH9.5.Theflaskswerethensealedandthesolutionswerestirred for24hwithamagneticbaratroomtemperature.After centrifu-gation,1mLofeachsolutionwaswithdrawnandwasdiluted1/6 withdistilledwater.Theabsorbanceoftheresultantsolutionswas
acquiredusingaSHIMADZUUV-1800spectrometerbetween200 and800nm.
2.8. Thermalconductivity
Thethermalconductivitywasmeasuredfrom parallelepiped-cutsamplesofhuminsfoams(minimumdimensions3×3×2cm) usingthetransientplanesourcemethodwithaHotDiskTPS2500 Sthermalconductivityanalyzerunderfixedconditionsof temper-atureandmoisturecontent(20◦Cand40%relativehumidity). 2.9. Thermalstabilityandmasslossanalyses
Thermalstabilityandmasslossofhuminsfoamswere investi-gatedbythermogravimetricanalysis(TGA)usingaMettlerToledo TGA/SDTD 851 with a microbalance accuracy of ±0.1g. The datawereanalyzedwithSTAR©software.Milledhumins foams (10–12mg)wereplaced ina 70Laluminapan and submitted todynamicthermal programs.Thethermal stability was evalu-atedbetween25and1000◦Cwithaheatingrateof10◦Cmin−1 (airflow=80mLmin−1),andwasreportedastheT10value, corre-spondingtothetemperatureatwhich10%ofthemasswaslost. Theresultswerethenaveragedovertworeplicates.
2.10. Watervaporadsorption
Water vapor sorption–desorption isotherms were obtained withaMicromeritics3Flexautomaticdevice.Priortoanalysis,the sampleswereoutgassedundersecondaryvacuumat110◦Cforat least72h.Thesorptionmeasurementswereperformedbydosing watervaporatrelativepressures(P/P0)rangingfrom1%to90%at 20◦C,andmeasuringtheequilibriumsorbedwaterforeachvalue ofP/P0.OnceP/P0=90%wasreached,desorptionwasmonitored, and thecorrespondingequilibrium sorbedwater volumeswere measuredtobuildthedesorptionbranch.
2.11. N2andCO2adsorption
Nitrogen(N2)andcarbondioxide(CO2)adsorptionisotherms at −196 and 0◦C, respectively, were obtained withASAP 2020 and ASAP 2420 (Micromeritics) automatic adsorption devices, respectively.Thesampleswereoutgassedundervacuumpriorto adsorptionanalysisforatleast48hat110◦C.
Using Microactive® and SAIEUS® software provided by Micromeritics,thefollowingparameterswerecalculated:(i)BET area, ABET (m2g−1), using the Brunauer–Emmet–Teller method [50];(ii)totalporevolume,V0.92,N2(cm
3g−1)orGurvitchvolume, takenasadsorbedvolumeattherelativeN2pressureof0.97;(iii) Dubininmicroporevolume,VDR-N2orVDR-CO2(cm3g−1),fromthe Dubinin–RaduskevichmethodappliedtotheN2orCO2isotherms, respectively;(iv)specificsurfaceareaand microporevolumeby applyingthedualgasanalysismethod(2D-NLDFT)methodtoN2 andCO2isotherms,SNLDFT(m2g−1)andV,NLDFT(cm3g−1), respec-tively[51,52].Animprovedversionofthe2D-NLDFTmethod,which takesintoaccountthediffusionlimitationinultramicropores,was applied[53].Theporesizedistribution(PSD)wasalsodetermined byapplyingthe2D-NLDFTtothecalculationof:ultramicropore volume(poresizelessthan0.7nm)V<0.7,NLDFT (cm3g−1), super-micropore volume(poresizebetween0.7and 2nm)V0.7–2,NLDFT (cm3g−1),microporevolumeV
micro,NLDFT(cm3g−1),totalpore vol-umeVT,NLDFT(cm3g−1),andaverageporevolumewav(nm). 2.12. CO2adsorptionaboveroomtemperature
22mgofmilledhuminsfoamswereplacedinaplatinum cru-ciblewithapiercedlid.ThesampleswereanalyzedwithTGAunder
agasflowof80mLmin−1usingthefollowingprogram:from25to 120◦Cat30◦Cmin−1underN2(necessarytoremoveallgases pre-viouslyadsorbed);isothermof30minat120◦CunderN2;cooling tothetemperatureofexperiment(25/50/70/100◦C)at30◦Cmin−1 underN2;equilibriumof30minunderN2.Atthispoint,thegas wasswitchedtoCO2andtheweightchangewasrecordedfor1h. TheadsorbedCO2isexpressedinwt.%relativetothemassofthe sample.ThesaturationoftheCO2adsorption,correspondingtothe pointwherenoadditionalCO2canbeadsorbedonthesurfaceof thehuminsfoamisobservedwheretheincreaseofthemassinthe thermographreachestheplateau,andisexpressedinminutes. 2.13. Tapdensitydetermination
TapandbulkdensityofF2900wasmeasuredwiththetapped densityanalyserAUTOTAP(QuantachromeInstruments)and fol-lowingtheASTMD8176-18standardtestmethod[54].
2.14. Preparationofactivatedcarbon
Approximately250mg of blockF2900sample wasweighted accuratelyandplacedinanaluminacrucible.Then,thesamplewas placedinatubularquartzfurnace.Thesamplewasheatedto900◦C underN2(60mLmin−1)at5◦Cmin−1.Then,thegaswasswitched toCO2(60mLmin−1)andheldat900◦Cfortheselectedactivation time(10,20,40or60min).Finally,thegaswasswitchedbackto N2,andthefurnacewasallowedtocooltoroomtemperature.
3. Resultsanddiscussion
3.1. Porousstructureandcomposition
Crudehuminsdirectlycollectedattheindustrialplantand with-outanypurificationormodificationstepcanbeauto-crosslinked andfoamedbyadirectthermalactivationprocess.However,the controlofboththeporosityandthemorphologyofthefinalfoam requirestheoptimizationofalltheparametersinvolved(heating
ramp,finaltemperatureofthetreatment,natureofthemold,etc.). Forinstance,usingaheatingrateof1◦Cmin−1inatubularoven undera flow of N2, nofoam isproduced but a completelyflat materialisobtainedinstead.Thechemicalreactionsandthegas evolutiondynamicsareindeedsoslowthatnotenoughbubbles areproduced,andthereremainsaftercuringaquitelowporosity inthematerial.Furthermore,ordinarytubularfurnacesusedfor carbonizationcanonlyoperatewithlimitedamountsofmaterial, whereasthenatureoftheavailablecruciblesisgenerallyrestricted toceramics.
Tobettercontrolporosityandfoamingyield,atwo-step prepa-rationissuggested.Thefirstheatingstepcanbecarriedoutinair inanykindofovenusingtemperaturesbelow350◦C.Underthese conditions,anytypeofmold,crucible,amountofhuminsand heat-ingrampcanbeeasilyappliedwithoutconstraints(e.g.1kgofcrude huminsinaluminumpans).Inthisway,itispossibletoproduce largeamountofhuminsfoamsinashorttime,whilethe tempera-turecanbeadjustedaccordingtothedesiredporosity.Afterwards, theas-preparedhuminsfoamscanbecarbonizedatanyselected temperatureunderN2and,astheyhavebeenthermosetinthe pre-viousstep,theheatingrateusedduringthecarbonizationdoesnot influencetheporosity.
Basedonthisprocedure,westudiedtheeffectofthis 2-step treatmentonindustrialhuminsandthemainresultsconcerningthe synthesisyieldandthecharacteristicsoftheprocessarepresented inTable1.ThesamplescalledF250hasbeenpreparedfromcrude huminsthrougha1-steptreatmentat250◦C(Fig.1),whileF2500 andF2900(2ndstep)havebeenpreparedfromF250by carboniza-tionat1◦Cmin−1from25to500◦Candto900◦C,respectively.
Onehourofisothermaldirecttreatmentat250◦Cresultedina finalfoamF250withavolume345%higherthanthatofthe start-ingcrudehumins.Duringthisprocess,amasslossofabout20%was observed,whichisslightlylowerthanthatreportedinourprevious work[55]duetoadifferentthermalprogram(here,adirect isother-maltreatmentinsteadofadynamicheatingramp).Thismethod wascarriedoutinthetemperaturerangecorrespondingtothefirst stepofmasslossdeducedfromtheTGAofhumins(140and260◦C).
Table1
Datarelatedtothepreparationofhuminsfoams,in1or2steps.TheyieldsofthelaststepindicatethepreparationyieldforF250(singlestep),andthecarbonizationyield forF2500andF2900.
Yieldoflaststep(wt.%) Totalyield(wt.%) Foamingcapacity(vol.%) Shrinkage(fromF250)(vol.%) Bulkdensity(gcm−3)
F250 78–80 78–80 345 – 0.055
F2500 50–52 41–43 290 17 0.070
F2900 45.5–47.5 36–37 265 23 0.092
Table2
TexturalcharacterizationfromN2andCO2adsorptiondataat−196and0◦C,respectively,forallmaterialstested.
N2 CO2 N2+CO2
ABET(m2/g) VDR-N2(cm3/g) V0.97(cm3/g) ABET(m2/g) VDR-CO2(cm3/g) SNLDFT(m2/g) V<0.7,NLDFT
(cm3/g) V0.7–2,NLDFT (cm3/g) Vmicro,NLDFT (cm3/g) VTotal,NLDFT (cm3/g) F250 – – 0.00 – – – – – – F2500 – – 0.01 224 0.18 493 0.12 0.02 0.14 0.14 F2700 – – 0.01 328 0.33 756 0.17 0.02 0.19 0.19 F2800 736 0.28 0.29 445 0.33 893 0.16 0.14 0.30 0.30 F2900 837 0.31 0.33 477 0.26 970 0.14 0.19 0.33 0.33 F21000 711 0.27 0.28 481 0.28 930 0.18 0.14 0.32 0.32
AccordingtoTGA-MSanalysis,itwasassociatedwiththereleaseof low-molecularweightspecies(mainlyH2O,CO2,COandMeOH) [55].Although TGA-MSwasunabletodetectanyother volatile compoundemittedbythematerialduringthermalcrosslinking, headspaceanalysisofthecrudehuminsevidenced thepresence ofotherspecies,especiallyfuraniccompounds.Head-space anal-yseswereperformedoncrudehuminsat170◦C(slightlybelow thehardeningtemperature), andare consistentwithpreviously reportedresults[38].Duringthisheatingprocess,16%ofthetotal masslossoftheuntreatedsamplewasfoundinTGA,corresponding tothereleaseofvolatilecompoundspresentinindustrialhumins. Thesespeciescontributemainlytotheemissionofgasesduringthe heattreatmentofhumins,leadingtotheformationofcellsinthe finalporousthermoset.Accordingtoheadspaceanalysis,ofthese 16%ofmassloss,onlyabout7.9%werefuraniccompounds,mainly unreactedmoleculesderivedfromtheACDprocessandtrappedin thematrix.
Accordingtothecharacterizationofthesurfacetexture(BETand NLDFT)indicatedinTable2,thefoamF250hasanegligiblesurface area,sincetheheattreatmentwastoomildtoproduceporosity. Incontrast,thefoamF2500,obtainedbycarbonizationunderN
2of F250at500◦C,presentedamasslossofapproximately50% dur-ingitspreparationandmadeitpossibletoobtainafinalyieldof about42%.Inthiscase,thefoamingcapacitywaslowerthanthat ofthestartingsampleF250,i.e.,withavolume290%higherthan thatoftheinitialcrudehumins.Thisisduetothermallyinduced rearrangementsofthepolymernetworkandthearomatizationof thestructureshownbytheFT-IRanalysis[55],whichledtoa17% shrinkageofF2500comparedtoF250,withoutmodifyingtheglobal morphology.Again,thesurfaceareaofthisfoamremainedlittle changedsincethetemperaturewastoolowtoproduceadditional porosity.
Asinthepreviouscase,thefoamF2900waspreparedby car-bonizationunderN2ofF250at900◦C.Duringthisstep,thematerial losesamassfractionabout55%higherthanthestartingF250,which givesafinalyieldofabout36%.Duetothehighertemperature,the shrinkagewashigherthanthatofF500(23%),givingafoamvolume 265%higherthanthatofthestartingcrudehumins.
IntermsofBETarea(ABET)huminsfoamsdonothavealarge surfacearea[56].However,whenthesurfaceareawascalculated bythe2D-NLDFTmethod,SNLDFT,assumingslitpores,theresult washigherthanexpected,reachingvaluesof756and970m2g−1 aftercarbonizationat500and900◦C,respectively.Thesurfacearea ofF2900isthusparticularlyhighcomparedtothatmeasuredfor tannin-derivedcarbonfoams,alwayslowerthan1m2g−1[57,58]. ThehighervalueofSNLDFTwhencomparedtoABETisduetothe exis-tenceofverynarrowporosity,whichisonlyaccessibletoCO2at 0◦Corinwhichonlyamonolayerofnitrogencanpenetrate.Inthis sense,theapplicationoftherefined2D-NLDFTmodel,developed fortheanalysisofbiochars[53],confirmedthatthebest simul-taneousfittingofN2 andCO2 isothermswasobtainedwhenthe lowerlimitofporewidth,wmin,oftheN2kernelwas4.6nmand not3.6nm.Theadjustablevalueofwmin hasaphysicalmeaning asit providesan estimateof theminimumporewidth
accessi-Fig.2.(a)SimultaneousfitofbothN2andCO2isothermsoftheF2900foambythe refined2D-NLDFT-HSmethod(thelinesrepresentthepredictionsofthemodel); and(b)pore-sizedistributionobtainedbyapplicationofthe2D-NLDFTmethod.
ble toN2 molecules inthepresent conditions ofmeasurement. Fig.2showsthegoodsimultaneousfitofN2andCO2isotherms andthecorrespondingpore-sizedistribution(PSD)obtainedfrom this fit. The PSD showedthat thecarbon foam wasexclusively microporous,withanaverageporediameterof0.78nm.ForF2900, morethanhalfofthemicroporositycorrespondedto ultramicro-pores,V0.7,NLDFT=0.16cm3g−1,whilethetotalmicroporevolume wasVmicro,NLDFT=0.30cm3g−1.Thetotalvolumecalculatedbythe 2-NLDFTmodelwashigherthanthatdeterminedbyN2adsorption, duetonitrogendiffusionlimitationsat−196◦C.Thisalsoagrees withthehigher micropore volume ofF2900determinedby DR methodfromCO2adsorptiondata,VDR-CO2,comparedtothat
deter-Table3
Elementalanalysisofhuminsfoams.
Sample C(wt.%) H(wt.%) O(wt.%) N(wt.%) F250 61.25 4.70 34.56 0.06 F2400 74.45 3.50 23.02 0.08 F2500 84.25 3.10 12.20 0.13 F2600 89.95 2.20 6.01 0.26 F2700 93.20 1.20 6.25 0.48 F2900 94.28 0.75 5.60 0.39 F21000 96.06 0.19 1.24 0.29
Fig.3. (a)VanKrevelendiagramand(b)evolutionofcarboncontentasafunctionoftemperatureforthecarbonizationofhuminsfoams.
minedfromN2adsorptiondata,VDR-N2.Therefore,thepresenceof
narrowultramicroporeswithporediametersoflessthan0.5nm, wherediffusionalresistancesofnitrogenat−196◦Careimportant,
makesSNLDFT>ABET,VDR-CO2>VDR-N2,andVT,NLDFT>V0.97.
Intermsofmacroporosity,accordingtoourobservations,the carbonizationstepsofF250at500◦C(inthepreparationofF2500)
and900◦C(inthepreparationofF2900)hadaminorimpactonthe
porousstructure,theshapeofthecellsshapeorthegeneralaspect ofthefoams.F250hadindeedahomogeneousanduniform macro-porosity,withporediametersbetween0.5and0.7mm,whilethe cellswereclosed(Fig.1).Thesepropertiesweremaintainedinthe derivedfoamsobtainedbycarbonizationat500and900◦C,except forthecorrespondingshrinkage,whichslightlyreducedthepore diameters(0.3–0.5mminF2900).Suchshrinkagedidnotalterthe shapeofthematerialbutonlyreduceditssize.Indeed,carbonizing F250parallelepipedsto500and900◦Cledtosmallerfinalsamples butmaintainingexactlythesameproportions.
Thebulkdensityofthefoamsincreaseswiththecarbonization temperature,from0.055gcm−3forF250,to0.070gcm−3forF2500, and0.092gcm−3forF2900(Table1).Thiseffectisduetoboththe shrinkage/aromatizationandthecorrespondingincreaseofcarbon content.
Indeed,intermsofelementalcontent,ascanbeseenfromthe elementalanalysis(Table3)andthevanKrevelendiagram(Fig.3), theincreasein thecarbonizationtemperatureof huminsfoams leadstoahigherandhighercarboncontent,whiletheoxygenis graduallyeliminated.FromaCcontentof53wt.%incrudehumins, itwaspossibletoreachabout90%ofCwithacarbonization tem-peratureof600◦C.AmaximumCcontentof96%wasreachedin caseofcarbonizationat1000◦C,whichcorrespondstoanalmost purecarbonmaterial.In thiscase,theresidualOisonlyaround 1.2wt.%.Thesecarbonfoamsareelectricallyconductivematerials thatcouldfindapplicationinseveralfields(e.g.,inthepreparation ofsacrificialelectrodes,electricaldevices,supercapacitators)oras microwaveabsorbersandshieldsforelectromagneticprotection.
Thewater adsorptiondata at25◦Cof materialsF250,F2500 andF2900areshowninFig.4.Thefirstpartofthewater adsorp-tion isotherm,i.e., at P/P0<0.3, reveals affinityof materials for water and is closely related to the nature and the amount of surfacefunctionalgroups.TheslopeoftheisothermatP/P0<0.3 decreasedintheorderF250>F2500>F2700>F2900,indicatingthe removalofoxygenatedgroupsatincreasingtemperature,in agree-mentwiththeelementalanalysis(Table3).Thesecondpartofthe wateradsorptionisothermisrelatedtothesurfacearea,andhence themaximumamountofadsorbedwaterdecreasedintheorder F2900>F2700>F2250>F2500,ingoodagreementwiththeresults obtainedbyN2 andCO2 adsorptionalreadyreportedinTable2. Themaximumwateruptakeof27.6wt.%correspondstoF2900at P/P0=0.92.
3.2. Surfacechemistry
Duringthefoamingprocessofhumins,theOcontentcanbe adjustedwiththeCcontent,whichprovidesagoodindicationofthe surfacereactivityofthematerial.AsshowninTable3,theOcontent decreasedfrom34.5wt.%inF250toabout12wt.%inF2500,and evento5.6wt.%inF2900.TheminimumwasfoundforF21000,with only1.2wt.%ofoxygen.Suchadecreaseintheamountofoxygen specieswasexpectedduringthecarbonizationstep.However,since someofthisoxygencontentcanbeinvolvedinresidualfuranrings orinmoietiesofthecrosslinkedstructure(suchasether,esteror ketonegroupsinthebulk),itisnotpossibletoevaluatethesurface behaviorofthesematerialsonlybasedonthisinformation.
Bystudyingtheoxygenatedfunctionspresentonthesurface ofthematerial,usefulinformationcanbeobtainedconcerningthe possibleapplicationsofhuminsfoams,suchasadsorbentsforwater treatmentorcatalystsupports. Inthis context,oneof themost importantparametersdescribingthevariablesurfacechargeisthe pointofzerocharge(PZC),whichisthepHthatanaqueous
solu-Fig.4. Wateradsorption–desorptionisothermsofhuminsfoamsF250,F2500,F2700 andF2900:(a)completedataset;(b)zoomontherangeoflowrelativepressures.
tioninwhichthematerialisimmersedmusthavetoproduceatotal surfacechargeequaltozero(neutrality)[59].Whenthematerialis immersedinasolutionwhosepHisabovethePZC,itssurfacehas anetnegativechargeandthereforehasacation-exchange capac-ity(CEC).Ifthesamematerialisimmersedinasolutionwhose pHisbelowitsPZC,ithasanetpositivechargewiththeabilityto retainanionselectrostatically,i.e.,ithasananion-exchange capac-ity(AEC).TodeterminethePZC,severalmethodscanbeused,which maycauseslightdifferences,dependingonthetechnique.
ThepHdrifttestisamethodgenerallyusedtoevaluatethePZC ofaseriesofsimilarmaterials[60,61].Inthismethod,acertain amountofmilledsampleisaddedtoanaqueoussolutionofknown pH,andthevariationofpHisrecordedsothatacurveoftheinitial pH(pHin)withrespecttothefinalpH(pHfin)canbebuilt.ThePZC isthusobtainedastheintersectionbetweentheresultantcurve pHinvs.pHfinandthelinepHin=pHfin.Anothermethodisthemass titrationmethod[62,63],initiallydevelopedforOH-richsurfaces. Themethodisbasedontheprinciplethat,byaddingmoreand moresolidtoasolutionofknownpH,thelatterwilltendtoward apHlimitvaluecorrespondingtothePZC.Therefore,startingfrom 3solutionswithdifferentinitialpH,andaddingtoeachofthem anincreasingamountofsolid,itispossibletobuild3pHcurves thattendtothesamevaluecorrespondingtothePZCofthesolid. However,thismethodhassomedisadvantages,becausetheexact valueassociatedtopHwithamassmaterialtendingtoinfinityis notalwayswelldefined(thecurvessometimestendtowardslightly differentpH),andthetechniqueistheoreticallylimitedtosamples freeofimpurities. However,we usedit torechecktheprevious valuesobtainedbythepHdriftmethod.Huminsfoamsprepared atlowtemperature(F180,F250andF300,respectivelypreparedat 180,250and300◦Cunderair)werealsotested.Despitethegood agreementfoundbetweenthetwotechniqueswhenusingthese foams,fluctuationswerenotedforF2500andF2900.
Table4
PZCobtainedfromdifferenttechniques.
Sample pHdriftmethod masstitration
F180 1.98 2.0
F250 2.0 2.0
F300 2.0 2.02
F2500 6.6 7.2
F2900 7.9 6.8
TheresultsobtainedarereportedinTable4.Theyshowthat F180,F250andF300foamshaveanetnegativeacidicsurface,with anotablevalueofPZCofabout2.Thismeansthatthesurfaceof thesefoamsisnegativewhenimmersedinwaterwithpHvalues above2,andispositiveonlyunderextremeacidicconditions,with awaterpHbelow2.Since,incommonaqueousapplications,thepH usuallyrangesbetween4and10,itcanbeassumedthatthesurfaces ofF180,F250andF300arealwaysnegativelycharged.Ontheother hand,thevaluesobtainedforF2500andF2900withpHdriftand masstitrationmethodsshowsomediscrepancies,butwithinthe limitofonepHunit.Wecanhoweverapproximateforboth sam-plesaPZCofaround7.Accordingtotheseresults,carbonization at500◦Cunderinertatmosphere(N2)isalreadyabletoproducea neutralsurfacebyremovingoxygen-basedmoietiesfromthe sur-face,inagreementwiththeelementalanalysis.Incontrast,thePZC obtainedforfoamstreatedatlowertemperatures(F180,F250and F300)inair(oxidizingconditions)areextremelypromisinginCEC applications.
Inordertobetterunderstandthenatureoftheremarkablylow PZCvaluesofF180,F250andF300,theBoehmtitrationmethod
[48,49,64,65] wasused ona milledF250 foam.The resultsare reportedinTable5.
Boehm’stitration methodis commonly usedtoidentifyand quantifyoxygen-basedfunctionalgroupsonthesurfaceofporous materials[49,64–68].Thetechniqueisbasedontheassumption that bases/acidsofgivenstrengths canonlyneutralizestronger acids/bases.Therefore,itisassumedthatNaHCO3(pKNaHCO3=6.37) canonlyneutralizecarboxylgroups,thatNa2CO3(pKNa2CO3=10.25) neutralizescarbonylsandlactones,whileNaOH(pKNaOH=15.74) neutralizesalltheacidicgroupsofthesurface(carboxyls,lactones andhydroxyls).Thetotalamountofbasicsurfacegroupscanbe evaluatedbytitration withastrongacidsuchasHCl.According totheresults,thesurfaceacidityofF250canbejustifiedbythe presenceofcarboxylicacidicgroupsonthesurface,accompanied bythetotalabsenceofbasicsites.Themainmoietiespresentare OH-groups,whilelactonesarepresentonlyintheformoftraces.
Since thePZCoffoams preparedbetween180◦C (minimum foamingtemperature)and300◦Cisthesame,itcanbeassumed thatthechemicalmoietiesonthesurfaceofthesematerialsare alsoverysimilar.ThesesurprisinglylowPZCvaluessuggesttheir potentialapplicationtodepollutionofwaterbyCECeffect. 3.3. Chemicalstability
Fromtheeconomical, practical andecological pointof view, themostinterestingfoamistheonethatcanbepreparedatthe lowesttemperature(F180).However,sincethechemicaland ther-malstabilityofsuchmaterialsisinfluencedbythetemperatureof preparationitself,wemustevaluatethisaspectinordertoidentify theminimumpreparationtemperatureoffoamsamplesthatcan beappliedincontactwithsolvents.
Toevaluatethechemicalstabilityofhuminsfoams,samplesof F180,F250andF300weremilledandwashedextensivelywith boil-ingdistilledwater,andthewashingswereanalyzedbyUPLCand GC–MS.WhileF250andF300washsolutionswereclearandno dis-solvedspeciesfoundinside,confirmingthegoodstabilityofthese
Table5
BoehmtitrationresultsforF250huminsfoam.
Sample ABET Carboxylicgroups Lactonicgroups Hydroxylgroups Basicgroups Totalacidicgroups
m2g−1 mmolg−1 mmolm−2 mmolg−1 mmolm−2 mmolg−1 mmolm−2 mmolg−1 mmolm−2 mmolg−1 mmolm−2
F250 0.2 0.2 0.9 0.09 0.3 0.4 1.6 0 0 0.7 2.8
Fig.5.UV–visabsorbancespectraoffoamF180washingsolutionsusingethanol (EtOH),acetonitrile(ACN),orwateratdifferentpH(4and9.5).
foams,theresultantsolutionfromF180hadanintenseyellowish color.FuranicspecieswerefoundinthissolutionbyUPLCanalysis, leadingtotheconclusionthatF180isnotchemicallystable.Thiscan beexplainedbyincompletecrosslinkingofthestructureat180◦C, leadingtothereleaseofmonomersorsmalloligomericspecies.
Additionalsolubilitytestswerecarriedoutbyadding0.1gof milledhuminsfoamsto15mLofdifferentsolutions(wateratpH 4,wateratpH9.5,EtOH,acetonitrile)and stirringtheobtained suspensions for24h. Again,F250and F300demonstrated good stability,sincenosolublespecies wasfoundinthe correspond-ingfilteredsolutions.Ontheotherhand,thepresenceoffuranics dissolved in F180 washing solutions was observed by UV–vis spectroscopy(Fig.5).180◦Cishereconfirmedtobeatoolow tem-peraturetoallowcomplete crosslinkingof humins.The species releasedcouldcomefromeither(i)smallmoleculesreleasedby hydrolysisfromthesurfaceofthefoams,or(ii)unboundedfuranic speciestrappedinthehuminsmixtureduringtheACDprocessof sugars.Furthermore,accordingtoheadspaceanalysis,the treat-mentof F180at250◦C causesalossof massof 8.5%,of which 4.2wt.%comesfromfuraniccompoundsreleasedin theformof volatiles.Thiscanfurtherjustifytheabsenceofthesecompoundsin F250,whereallvolatilesarereleasedatahigherfoampreparation temperature.Accordingtotheseresults,itisnotrecommendedto useF180inapplicationsinvolvingimmersioninsolutions. How-ever,F250hasshown goodchemical stability,sotheminimum temperatureforpreparingastablefoamforsuchapplicationscan besetat250◦C.
3.4. Thermalbehavior
Thethermalconductivity,thebulkdensity()andthevalueof T10,i.e.,thetemperatureatwhichthematerialloses10%ofitsmass, ofseveralhuminsfoamsarereportedinFig.6.Itcanbeseenthat F180hasamuchhigherdensity(0.132gcm−3)thanthatofother huminsfoams(below0.1gcm−3).Indeed,180◦Cdoesnotprovide asufficientmassloss(only5.5wt.%witharampof10◦Cmin−1),so thatlittlegasisemittedduringthepreparation,whichleadstoa
Fig.6.Thermalconductivityand T10 values(derivedfromTGAat10◦Cmin−1 between25and1000◦Cunder50mLmin−1airflow)asafunctionofthebulkdensity ()of5samplesofhuminsfoam.
lowerporosityinthefinalmaterial.Byusinghighertemperatures of220◦C(masslossof12wt.%)and250◦C(masslossof20wt.%), foamsF220andF250areproducedrespectivelywithlowerdensity andlowerthermalconductivity.Afurtherincreaseinthe prepara-tiontemperatureresultsinanincreaseinthedensityofthefoams duetoshrinkage,aswellasanincreaseinthethermalconductivity duetoaromatizationandincreasedCcontent.Themaximumvalue ofthermalconductivitywasfoundforF2900,0.070Wm−1K−1,as predictedbyitselementalcomposition(almostpurecarbon)and itshigherdensity(0.092gcm3).Afteroptimization,foamsF220, F250andF2500mightfindinterestingapplicationsasthermal insu-latingmaterials.Indeed,accordingtotheFederationofEuropean RigidPolyurethaneFoamsAssociations,amaterialcanbe consid-eredasgoodinsulatorifitsthermalconductivityisintherange 0.035–0.055Wm−1K−1[69].
Ontheotherhand,theT10valuecanbeusedtoevaluatethe thermalstabilityofhuminsfoams.Inunreactedcrudehumins,T10 wasfoundtobearound200◦C[55],whereasthisvalueishigher inhuminsfoams.Withregardtothelatter,T10increaseswiththe preparationtemperature.Thisresultwasexpected,sinceoncea giventemperatureisreachedduringthepreparationofthefoam, allthepossiblereactionsthatcanoccuratthistemperatureare performed.Newreactions,andthereforenewlossofmass,canbe observedonlywhenthetemperatureisfurtherincreased.Thus,for instance,whentheF300foamispreparedat300◦C,weassumethat alltheprocessesthatcouldhavehappenedatalowertemperature (e.g.100,150,200,250◦C)havealreadyoccurred,andthatnomass losscanbeobservedbelow300◦C.
The maximum value of T10 was reached with F2500 (T10=500◦C),which correspondstothevalueatwhich thermo-oxidativedegradation(pyrolysis)occurscompletelyinairunder theconditionsusedhere.Above500◦C,thematerialburnt com-pletely.Thus,underair,nohigherthermalstabilitycanbeachieved forhuminsfoamsusingdynamicheatingof10◦Cmin−1,asusedin thiswork.
To further evaluate the thermal stability of humins foams, headspace analyses of F180, F250 and F300 were carried out. Althoughtheseresultswereobtainedbyaqualitativeanalysis,we
attemptedtogetsemi-quantitativedatausingthetotalmassloss ofthesampleobtainedbyTGAandthecomparisonsbetweenthe areasofsignalspeaksinGC–MS.Theseresultsprovideadditional informationaboutthenatureofthegasesemittedwhenhumins foamsaresubjectedtoheatingprocessesorthermalsources.
F180alreadyreleasesgasesat100◦C,withamasslossof0.5%.Of thesegases,about70%(correspondingtoabout0.35%ofthetotal masslostbyF180at100◦C)correspondstofuraniccompounds, whileabout23.6%,correspondingto0.12%ofthetotalmassofF180, isassociatedwithdegradationcompounds(i.e.,acetone,MeOHand aceticacid).
ByincreasingtheF180treatmenttemperature,newunknown signals were foundin the chromatogram.A possible candidate might be 5-methoxyfuran-2-carbaldehyde, not present in our library.Asexpected,thehighesttemperatureinvestigated(250◦C) gavethelargestamountofgasemittedbyF180,whichcorresponds inanycasetoonly8.5%ofthetotalmassofF180.Themaximum sig-nalforF180at250◦Cisassociatedwithfuraniccompounds,which correspondtoonlyabout4.21%ofthemassreleasedbyF180in2h. Specifically,andaccordingtoTGA,thismasslossoccurswithinthe first45min,afterwhichnofurthermasslossisdetected, regard-lessofthedurationofthethermaltreatment.Therefore,wecan concludethatnohazardsareassociatedwithF180duringsimilar thermalevents.
As expected, the other foams studied (prepared at higher temperatures)showedgreaterthermalstability.Nospecieswas emittedanddetectedbytreatmentofF250andF300to100and 150◦C.At200◦C,F250hadamasslossoflessthan1%,mainly asso-ciatedwithdegradationproducts.Thelossofmassdetectedfor F300wasevenlower,andnofuranicscouldbeidentified, suggest-ingtheircompletereleaseduringthefoampreparationat300◦C. HeadspaceanalysisofF300at200◦Cidentifiedwaterasthemain releasedcompound,whichcaneitherbearesidualby-productof condensationreactionsororiginatefromwaterformerlyadsorbed fromtheenvironment.Aceticacidwasthemaindegradation prod-uctfoundforF300at250◦C.Atthesametemperature,tracesof 2,3-butadieneand2-butanonealsoappearedasdegradation prod-ucts,possiblyassociatedwiththehighercarboncontentofF300. Thesevaluesaremainlyrelevantfordescribingtrends.However, consideringthesevaluesbeingtheresultofanoverestimation,we envisagetheworstpossiblescenarioofevolutionofvolatiles.In thecasewherehuminsfoamswouldbeaccidentallysubjectedto thermalsources,wecanforeseeanextremelylimitedriskforthe environmentandhumanhealth,inagreementwiththepreviously reportedstudies[70,71].
3.5. CO2adsorption
TheapplicationofhuminsfoamstotheadsorptionofCO2ata temperaturehigherthanorequaltoroomtemperaturehasbeen tested,andtheresultsarepresentedinTable6.CO2isamolecule thatcanbeeasilyadsorbedinmaterialshavingmicro-and ultrami-cropores.Duetoitslowsurfacearea(Table2),F250,hastheworst performanceintermsofCO2adsorption.F2500andF2700adsorb moreCO2,andinafasterway.Thefastestadsorptionwasobserved forF2500,wheresaturationwasobservedafteronly37minat25◦C, whilethelargestamountofadsorbedCO2 wasfoundforF2700, with3.66wt.%ofCO2adsorbed.Byincreasingthetemperature,the adsorbedmoleculesarelessandlessretainedintheporous mate-rials.F2900presentedaquitehighadsorptioncapacity,although lessthanthepreviouscases,itsbestvaluebeing2.64wt.%ofCO2 adsorbedat25◦C.Theseresultsarecomparablewiththosereported formaterial studiedfor similarapplication,withtheadvantage thathuminsfoamsarecompletelybioderivedandthatnosurface activationhavebeenperformed[72].
Table6
CO2adsorptioncapacityofhuminsfoamsF250,F2500,F2700andF2900atdifferent temperatures(25,50,70and100◦C).Thesaturationofthesurface,expressedin time,isfoundwhenonereachestheplateauofthethermogramunderflowofCO2 gas.
Temperature CO2adsorbed(wt.%) Saturation(min)
F250 25◦C/298K 0.52 >60 50◦C/323K 0.38 >60 75◦C/348K 0.25 >60 100◦C/373K 0.17 46 F2500 25◦C/298K 2.61 37 50◦C/323K 1.47 25 75◦C/348K 0.80 19 100◦C/373K 0.43 14 F2700 25◦C/298K 3.66 44 50◦C/323K 3.10 42 75◦C/348K 2.02 36 100◦C/373K 1.23 29 F2900 25◦C/298K 2.64 >60 50◦C/323K 2.56 >60 75◦C/348K 1.77 >60 100◦C/373K 1.17 >60 Table7
TapdensityobtainedwithAutotapanalyzer.
F2900(AUTOTAP)
Mass 4.849g
Volume 4.85mL Tapdensity 1.00gcm−3
Furthermore, we also found an almost perfect correlation betweenthedifferencesinCO2 adsorptionat50and100◦Cand
thevolumesofmicropores(<1nm)calculatedbyNLDFT(SI1and SI2).ThisconfirmsthatCO2adsorptioninvolvesmainlyporeswith
volumesminorthan1nm.
Thecapacityof storageathightemperatures(100◦C),which is relevantforcombustiongases,canbecompared inliterature withresultsforsimilarmaterials.Relevantdatahavebeenrecently reported about activated anthracite MSP-20X of KANSAI Coke (Japan)[73].Comparedtohuminsfoams,thelattercarbonmaterial has a higher surface area (ABET=2363m2g−1). Its CO2 capac-ity passedfrom 2.6mmolg−1 (11.7%) at25◦C to0.58mmolg−1 (2.55%) at 100◦C, not sofarfrom theresultshereobtainedfor huminsfoams.However,huminsfoamshavetheadvantagetobe bioderivedandsustainable,andthatnoactivationisperformed, whileMSP-20XisproducedfromanthraciteactivatedwithKOH.
ThehuminsfoamscouldbecompetitiveforCO2captureat rel-evanttemperatures,evenmoreiftheapparentdensityofthebed isconsidered.Tappeddensityofa powderiscalculatedasratio betweenthemassofthematerialandthevolumeitoccupiesafter thathasbeentapped.
Tap density=tappedmassvolume(g/mL)
Tapdensitycanbeseenastherandomdensepacking.Materials withhighsurfaceareagenerallyhavealowtapdensity,andsothe neededadsorptionvolume(columnortank)ishigher.ForMSP-20X ithasbeencalculatedatapdensityof0.35gcm−3,whilewefounda muchhighertapdensityof1.00gcm−3(Table7)ofgroundedF2900 huminsfoam.EvenconsideringthattheMSP-20XofKANSAICoke adsorbs2.55wt.%ofCO2at100◦C,comparedto1.17%forF2900we willneedatallercolumnofMSP-20Xbecausethebeddensityis muchlower(2.9times).Therefore,huminsfoamscancompetein suchapplications.
Table8
YieldandmorphologicalcharacterizationofactivatedcarbonmonolithspreparedfromF2900huminsfoamat900◦C,usingCO2flow=60mL/minfor10,20,40and60min. Time(min) Burnoff(%) SBET(m2/g) SNLDFT(m2/g) VDR-N2(cm3/g) VDR-CO2(cm3/g) VNLDFT(cm3/g) V0.97(cm3/g) Vmeso(cm3/g)
10 14.2 985 1151 0.37 0.33 0.40 0.41 0.01(3%) 20 36 1347 1326 0.49 0.34 0.52 0.56 0.04(8%) 40 94.8 1482 1289 0.52 0.30 0.56 0.68 0.12(19%)
60 100 – – – – – – –
Fig.7. RepresentationofCO2adsorptiontrendatdifferenttemperaturesforF250,
F2500,F2700andF2900.
3.6. Activatedcarbonmonolithspreparation
HuminsfoamF2900hasbeenusedinthepreparationof
acti-vatedcarbonwiththeCO2activationtechnique.Table8presents
theyieldsandthemaintexturalparametersobtainedbynitrogen andcarbondioxideadsorption,at−196and0◦C,respectively,for allthematerialstested.Theburn-off(lossofmaterialmass)isone ofthemostimportantparameters tocontrol inthepreparation ofactivatedcarbon.AsreportedbyMolina-Sabioetal.,inorder tomaximizetheporosityofthematerial,theburn-offshouldbe lessthan40%[74].Abovethisamount,althoughthematerialisstill consumedbyCO2,nohighersurfaceareaisreachedand,onthe con-trary,thesurfaceareaisincreasinglyreduced.Thisoccursbecause theporesformedduringtheearlystagesofactivationincreasein volumeandnumber,andwidentomeltineachother.This,along theirexternalburning,reducesthesurfaceareaobtainedduring thefirstactivationstep(Fig.7).
Fig.8. Activatedcarbonmonolithsproducedwith20and40minofCO2treatment at900◦C.
Table9
Elementalanalysisofactivatedcarbonmonoliths.Thedataareobtainedasaverages oftworeplicates.
Sampleactivation C/H %N %C %H %O 10min 61.46 0.12 90.61 1.47 8.15 20min 53.72 0.07 86.76 1.61 7.90 40min 72.96 0.07 76.18 1.05 14.31
Thebestresultwasobtainedusinganactivationtimeof20min (Fig.8),withaburn-offof36%andaBETsurfaceareaof1347m2g−1. Increasingtheactivationtimeto40minresultsinanalmost com-pletelossofmaterial,withayieldofonly5.2%,whileafter60min theentiresampleis completelyconsumed(Fig.8).The elemen-talanalysisreportedinTable9alsoshowshowtheCO2activation
Fig.10.(a)Poresizedistributionand(b)cumulatedporevolumecalculatedby NLDFT.
increasestheO%attheexpenseofC%.Thiseffectisduetothe factthatduringactivation,CO2reactswiththecarbonofthe mate-rial,consumingit(CO2+C→2CO).Asaresult,Cisprogressively consumed,andtheO/Cratiotendstoincrease.
F2900usedinthesepreparationshadanOcontentof5.60%and aCcontentof94.28%.After10and20minutesofactivation,the O%increasedbyabout30%whiletheC%decreasedby8%.This effectisevenmorepronouncedafter40minofactivation,whereO %reaches14.31%.
TheBETareaincreasedwithactivationtimeaswellasthe poros-ityintherangeofmicroandmesopores.Fig.9showstheadsorption isotherms of nitrogen (Fig. 9a) and carbon dioxide adsorption (Fig.9b)onnon-activatedF2900foamandactivatedcarbonfoams. By increasing the activation time, the elbow of nitrogen isothermswaswidened,indicatingabroadeningofthe microporos-ityandthetotaladsorbedvolumewasalsohigher.After40minof activation,theplateauwasnolongerhorizontal,butaslightslope appeared,indicatingtheexistenceofmesoporosity(poreswider than2nm)aswellasmicroporosity(poresnarrowerthan2nm). TheCO2 adsorptionisothermsgiveanindicationofthevolume ofporesnarrowerthan1.2nm.Activationtimesofupto20min creatednewpores,narrowerthan1.2nm, buthigheractivation timessimplyproducedporewidening.Thus,theadsorptionofCO2 decreasedand thatof N2 adsorptioncontinuedtoincrease.The latterwasconfirmedbytheporesizedistribution(PSD)andthe cumulativeporevolume(Fig.10)calculatedbytheNLDFTmodel.
4. Conclusions
Inthiswork,newmaterialscalledhuminsfoamshavebeen pre-paredfromhighlyviscoushuminsderivedfromtheindustrialacid catalyzeddehydrationoffructoseintomethoxymethylfurfural.
Thehuminsfoamswerecharacterizedintermsofcomposition, surfacefunctionality,porosity,andthermal andchemical stabil-ity. Thecarbon content of humins foamscan becontrolledvia thesynthesis protocol,andcan beashighas95.4%in thecase ofcarbonizationat900◦C.Thismakescarbonizedhuminsfoams interestingforapplicationssuchaselectrodesforenergystorage andconversion,ascatalystsupports,orascheapadsorbents.
Thesurfacechemistryofseveralhuminsfoamshasbeen inves-tigated.Foamspreparedatatemperatureof180◦Cexhibitpoor chemicalstability,releasingfuraniccompoundsinsolution.These furanicsmightbeunboundspeciespresentinthemixtureofcrude huminsand originatefromanincompletecrosslinking,orcome fromthehydrolysisofthehuminsfoamsstructure.Totalsolvent stabilitywasobtainedwithfoamspreparedattemperaturesabove 250◦C.Ofthefoamstested,thosepreparedunderairata tempera-turebelow300◦Chadanetnegativesurfacechargeinwater,with animpressivevalueofpHofzeropointchargeofabout2(while commercialproductsrangebetween4and10.5).Thankstothis low valueofpHofzero pointchangeassociated witha cation-exchangecapacity,thesehuminsfoamscanbeappliedinseveral fields(e.g.waterdepollution).Inaddition,thankstotheirrigidand crosslinkednature,theycanbepreparedinblocks ofindefinite dimensions(limitedonlybythedimensionsofthecrucibleand theoven)andusedasawhole.Thisaspect,aswellasthecost effi-ciencyofthecrudehumins,mightbeusefulforpreparingmaterials forthedecontaminationoflargevolumesofsolvents,forwhicha highsurfaceareaisrequired.Incontrastwithcommercially avail-ablecarbon-materialsusedfordecontamination(mainlybasedon modifiedactivatedcarbon)theseresultsarereferredtothebare carbonsupport,whereanyfurtheractivationstrategywasapplied. In ordertoenhancethematerialsproperties andmake humins foamsusefulin theadsorption ofcationic contaminant,further studiesshouldbeconducted,aimedtotheenhancementofthe sur-facearea(whereforcommercialproductsrangesbetween500and 2000m2g−1)andincreasingthedensityofchargeonthesurface. Treatmentswithacids,basesoroxidizingagents areoftenused formodifyingchemicalandphysicalpropertiesofcarbon mate-rials,and shouldbetestedandscreenedalsoforhuminsfoams. Again,thebigadvantageofhuminsfoamsisthethree-dimensional starting structure, that can be maintained in the final treated material, while commercial products are generally in form of powder.
Humins foams also have high thermal resistance, which increases with the temperature of preparation. According to headspaces analysis,all thefoamstested released onlya small amountofdegradation gasovera period of2h, thusexcluding therisksof toxicityin theeventof accidentalpresence ofheat sources. Themaximumamountof gasreleasedwasabout8.5% of thetotal massof thesample,obtainedby heattreatmentof F180 for 2hat 250◦C. F250 presented a higher thermal resis-tance,withonly0.43%ofmasslossover2h.Therefore,accordingto thedatareportedhere,huminsfoamspreparedatatemperature above250◦Cexhibitthehighestthermal andchemicalstability, unlikethosepreparedatalowercrosslinkingtemperature.F250 alsohasthelowestthermalconductivity(0.04Wm−1K−1). Consid-eringcommercialthermalinsulantssuchastherigidpolyurethane foams,whichshowlowerthermalconductivity(18–28Wm−1K−1) andalsolowerthermalresistancecomparedtohuminsfoams,it pavesthewayfornewinvestigationsasinsulationmaterial pro-videdoffurtheradjustment.Intheseterms,bychangingthenature ofthegasin thehumins foamscells,we canincreasethe
elas-ticityofthematerialorfurtherreducetheporessize(byfoams preparationunderpressurizedenvironment)askeystepsforfuture studies
Humins foamshave also been tested for CO2 adsorption at temperatures ≥25◦C. The best results have been obtained at 25◦C, whereastheefficiencydecreaseswiththeincreaseofthe adsorptiontemperaturetested.Huminsfoamscarbonizedat500◦C showedthefastest adsorptionrateat25◦C, reachingsaturation with2.61wt.%ofCO2fixedtothesurfaceafteronly37minof treat-ment.Moreover,thelargestamountofCO2 adsorbed,3.66wt.%, wasobservedat25◦Cforfoamscarbonizedat700◦C,whilethe saturation was reachedafter 44min of treatment. Considering the lack of similar studies in literature, further investigations should be performed, while a new application field could be explored.
Finally, the preparation of activatedcarbon monoliths from huminsfoamswastested. Physical(CO2)activationwascarried outat900◦C,reachingaBETsurfaceareaofupto1482m2g−1.The bestresultswereobtainedafter20minoftreatment,reachinga surfaceareaof1347m2g−1withonly36%burn-off.Theseresults arewaybetterthanthosereportedinliteratureforpreparations from humins, and are more time and cost effective. Further-more, the surface area obtained is competitive with those of activatedcarboncommerciallyavailable.Thetextural characteri-zationshowedthepresenceofmeso-andmicroporosity,aswell asporesnarrowerthan1.2nm.Theseactivatedcarbonscan com-pete withthose commercially available, withthe advantage of greatadaptability,becausetheycanbeproduced intheformof monoliths (easier tohandle andrecover), andwith a favorable cost.
We can assume that all theresults reported in this article, referringspecificallytohuminsfoams,couldbeextended (with thenecessarylimitations)toanymaterialbasedonthermosetting huminsandcomposites,whereviscouscrudehuminsareusedas startingmaterialandcrosslinked.
Funding
ThisworkhasbeenfundedtheEuropeanCommissionforits financialsupport:H2020MSCAproject“HUGS”,GA675325.
Authors’contribution
ThemanuscriptwaswrittenbyPTandAMandwasrevisedby alltheauthors.PTandCDSdidthefoamingpreparingexperiments. AlltheotherexperimentswereperformedbyPTandanalyzedby AM.GvK,CH,CL,AC,VF,AMcontributeininvestigation, concep-tualization,designofmethodologyandvalidationofresults.AM supervisedandobtainedthefunding.Allauthorshavegiventheir approvaltothefinalversionofthemanuscript.
Conflictofinterest
Nonedeclared.
Acknowledgements
SpecialthanksgotoPhilippeGadonneixforhishelpand assis-tancewithelementalanalysisandpotentiometrictitration,andto BlagojKarakashovforthehelpwiththermalconductivityanalysis.
AppendixA. Supplementarydata
Supplementarydataassociatedwiththisarticlecanbefound,in theonlineversion,athttps://doi.org/10.1016/j.apmt.2020.100622.
References
[1]S.K.R.Patil,C.R.F.Lund,EnergyFuels25(2011)4745–4755.
[2]I.VanZandvoort,Y.Wang,C.B.Rasrendra,E.R.H.VanEck,P.C.A.Bruijnincx, H.J.Heeres,B.M.Weckhuysen,ChemSusChem6(2013)1745–1758.
[3]J.Heltzel,S.K.R.Patil,C.R.F.Lund,in:Springer(Ed.),React.PathwaysMech. ThermocatalyticBiomassConvers.II,2016,pp.105–118,
Singapore.
[4]Z.Cheng,J.Everhart,G.Tsilomelekis,V.Nikolakis,B.Saha,D.Vlachos,Green Chem.20(2018)997–1006.
[5]I.VanZandvoort,E.J.Koers,M.Weingarth,P.C.A.Bruijnincx,M.Baldus,B.M. Weckhuysen,GreenChem.(2015)4383–4392.
[6]J.Herzfeld,D.Rand,Y.Matsuki,E.Daviso,M.Mak-jurkauskas,I.Mamajanov,J. Phys.Chem.(2011)5741–5745.
[8]G.R.Gomes,D.S.Rampon,L.P.Ramos,Appl.Catal.AGen.545(2017)127–133.
[9]C.Sievers,I.Musin,T.Marzialetti,M.B.V.Olarte,P.K.Agrawal,C.W.Jones, ChemSusChem2(2009)665–671.
[10]S.G.Wettstein,D.M.Alonso,Y.Chong,J.A.Dumesic,EnergyEnviron.Sci.5 (2012)8199–8203.
[11]Y.Román-Leshkov,C.J.Barrett,Z.Y.Liu,J.A.Dumesic,Nature447(2007) 982–985.
[12]R.M.Musau,R.M.Munavu,Biomass13(1987)67–74.
[13]C.Moreau,M.N.Belgacem,A.Gandini,Top.Catal.27(2004)11–30.
[14]X.Fang,Z.Wang,B.Yuan,W.Song,S.Li,W.Lin,ChemistrySelect3(2018) 12243–12249.
[15]D.W.Rackemann,W.O.Doherty,BiofuelsBioprod.Biorefining5(2011) 198–214.
[16]F.S.Asghari,H.Yoshida,Ind.Eng.Chem.Res.46(2007)7703–7710.
[17]M.-M.Titirici,R.J.White,N.Brun,V.L.Budarin,D.S.Su,F.delMonte,J.H.Clark, M.J.MacLachlan,Chem.Soc.Rev.44(2015)250–290.
[18]J.Shen,C.E.Wyman,AIChEJ.58(2012)236–246.
[19]B.Girisuta,B.Danon,R.Manurung,L.P.B.M.Janssen,H.J.Heeres,Bioresour. Technol.99(2008)8367–8375.
[20]R.Weingarten,J.Cho,R.Xing,W.C.Conner,G.W.Huber,ChemSusChem5 (2012)1280–1290.
[21]B.F.M.Kuster,Carbohydr.Res.54(1977)177–183.
[22]B.Girisuta,L.P.B.M.Janssen,H.J.Heeres,Ind.Eng.Chem.Res.46(2007) 1696–1708.
[23]L.Faba,D.Garcés,E.Díaz,S.Ordó ˜nez,ChemSusChem(2019)1–10.
[24]S.Xu,D.Pan,Y.Wu,N.Xu,H.Yang,L.Gao,W.Li,G.Xiao,Ind.Eng.Chem.Res. 58(2019)9276–9285.
[25]Z.Cao,Z.Fan,Y.Chen,M.Li,T.Shen,C.Zhu,H.Ying,Appl.Catal.BEnviron. 244(2019)170–177.
[26]M.R.Whitaker,A.Parulkar,P.Ranadive,R.Joshi,N.A.Brunelli,ChemSusChem 12(2019)2211–2219.
[27]D.Garcés,L.Faba,E.Díaz,S.Ordó ˜nez,ChemSusChem12(2019)924–934.
[28]P.Ramesh,A.Kritikos,G.Tsilomelekis,React.Chem.Eng.4(2019)273–277.
[29]S.Liu,K.Wang,H.Yu,B.Li,S.Yu,Sci.Rep.9(2019)1–9.
[30]A.Mija,J.C.VanDerWaal,J.Pin,N.Guigo,E.DeJong,Constr.Build.Mater.139 (2016)594–601.
[31]L.Filiciotto,A.M.Balu,A.A.Romero,E.Rodríguez-Castellón,J.C.vanderWaal, R.Luque,GreenChem.19(2017)4423–4434.
[32]S.Kang,G.Zhang,Q.Yang,J.Tu,X.Guo,F.G.F.Qin,Y.Xu,BioResources11 (2016)9496–9505.
[33]S.Kang,J.Fu,G.Zhang,W.Zhang,H.Yin,Y.Xu,Polymers(Basel)9(2017)373.
[34]A.Mija,J.C.vanderWaal,E.deJong,G.P.M.vanKlink,Processforthe modificationofhumins,WO2018/062995A1(2018).
[35]A.Schweizer,(1940)781–784.
[36]T.M.C.Hoang,E.R.H.vanEck,W.P.Bula,J.G.E.Gardeniers,L.Lefferts,K.Seshan, GreenChem.17(2015)959–972.
[37]I.vanZandvoort,TowardstheValorizationofHuminBy-Products: Characterization,UtrechtUniversity,2015.
[38]A.Sangregorio,N.Guigo,J.C.vanderWaal,N.Sbirrazzuoli,ChemSusChem (2018).
[39]X.Montané,R.Dinu,A.Mija,Molecules24(2019)4110.
[40]A.Mija,J.C.vanderWaal,G.vanKlink,E.deJong,Humins-containingfoam, WO2017074183A8(2016).
[41]A.Mija,E.deJong,J.C.vanderWaal,G.vanKlink,Huminscontainingfoam, WO2017074183A120170504(2017).
[42]J.M.Pin,N.Guigo,A.Mija,L.Vincent,N.Sbirrazzuoli,J.C.vanderWaal,E.de Jong,ACSSustain.Chem.Eng.2(2014)2182–2190.
[43]P.Tosi,G.P.vanKlink,A.Celzard,V.Fierro,L.Vincent,E.deJong,A.Mija, ChemSusChem11(2018)2797–2809.
[44]K.Kadirvelu,M.Kavipriya,C.Karthika,M.Radhika,N.Vennilamani,S. Pattabhi,Bioresour.Technol.87(2003)129–132.
[45]N.MohamadNor,L.C.Lau,K.T.Lee,A.R.Mohamed,J.Environ.Chem.Eng.1 (2013)658–666.
[46]S.Kang,J.Fu,Z.Deng,S.Jiang,G.Zhong,Y.Xu,J.Guo,J.Zhou,Sustainability10 (2018)16–19.
[47]D.Chernysheva,Y.Chus,V.Klushin,T.Lastovina,L.Pudova,N.Smirnova,O. Kravchenko,V.Chernyshev,V.P.Ananikov,ChemSusChem11(2018) 3599–3608.
[48]H.-P.Boehm,E.Diehl,W.Heck,R.Sappok,Angew.Chem.Int.Ed.English3 (1964)669–677.
[50]S.Brunauer,P.H.Emmett,E.Teller,J.Am.Chem.Soc.60(1936)309–319.
[51]J.Jagiello,C.Ania,J.Parra,C.Cook,CarbonN.Y.91(2015)330–337.
[52]J.Jagiello,J.P.Olivier,Adsorption19(2013)777–783.
[53]J.Jagiello,J.Kenvin,A.Celzard,V.Fierro,CarbonN.Y.144(2009)206–215.
[54]G.Sdanghi,V.Nicolas,K.Mozet,G.Maranzana,A.Celzard,V.Fierro,Int.J. HydrogenEnergy44(2019)16811–16823.
[55]P.Tosi,G.P.M.vanKlink,A.Celzard,V.Fierro,L.Vincent,E.deJong,A.Mija, ChemSusChem11(2018)2797–2809.
[56]J.Encalada,K.Savaram,N.A.Travlou,W.Li,Q.Li,C.Delgado-Sánchez,V. Fierro,A.Celzard,H.He,T.J.Bandosz,ACSCatal.7(2017)7466–7478.
[57]G.Tondi,V.Fierro,A.Pizzi,A.Celzard,CarbonN.Y.47(2009)1480–1492.
[58]W.Zhao,A.Pizzi,V.Fierro,G.Du,A.Celzard,Mater.Chem.Phys.122(2010) 175–182.
[59]F.I.Morais,A.L.Page,L.J.Lund,SoilSci.Soc.Am.J.40(1976)521–527.
[60]J.Rivera-Utrilla,I.Bautista-Toledo,M.A.Ferro-Garca,C.Moreno-Castilla,J. Chem.Technol.Biotechnol.76(2001)1209–1215.
[61]P.C.C.Faria,J.J.M.Órfão,M.F.R.Pereira,WaterRes.38(2004)2043–2052.
[62]J.S.Noh,J.A.Schwarz,J.ColloidInterfaceSci.130(1989)157–164.
[63]S. ˇZalac,N.Kallay,J.ColloidInterfaceSci.149(1992)233–240.
[64]H.-P.H.Boehm,CarbonN.Y.40(2002)145–149.
[65]H.P.Boehm,Adv.Catal.16(1966)179–274.
[66]A.Contescu,C.Contescu,K.Putyera,J.A.Schwarz,CarbonN.Y.35(1997) 83–94.
[67]S.L.Goertzen,K.D.Thériault,A.M.Oickle,A.C.Tarasuk,H.A.Andreas,CarbonN. Y.48(2010)1252–1261.
[68]I.I.Salame,T.J.Bandosz,J.ColloidInterfaceSci.240(2001)252–258.
[69]Bing,in:F.ofE.R.P.F.Associations(Ed.),Rep.N1,FederationofEuropeanRigid PolyurethaneFoamAssociations,2006,pp.1–33.
[70]A.Muralidhara,A.R.Division,I.National,D.Environnement,BiofuelsBioprod. Biorefining(2019)464–470.
[71]A.Muralidhara,P.Tosi,A.Mija,N.Sbirrazzuoli,C.Len,V.Engelen,E.deJong, G.Marlair,ACSSustain.Chem.Eng.(2018)16692–16701.
[72]E.S.Sanz-Pérez,M.Olivares-Marín,A.Arencibia,R.Sanz,G.Calleja,M.M. Maroto-Valer,Int.J.Greenh.GasControl17(2013)366–375.
[73]V.Fierro,E.García-Díez,S.Schaefer,A.Sanchez-Sanchez,M.M.Maroto-Valer, S.García,A.Celzard,Carbon2019,Lexington,USA,2019.
[74]M.Molina-Sabio,M.T.González,F.Rodriguez-Reinoso,A. Sepúlveda-Escribano,CarbonN.Y.34(1996)505–509.
